Undigestible, insoluble food particles, such as wheat bran, are important dietary constituents that serve as a fermentation substrate for the human gut microbiota. The first step in wheat bran fermentation involves the poorly studied solubilization of fibers from the complex insoluble wheat bran structure. Attachment of bacteria has been suggested to promote the efficient hydrolysis of insoluble substrates, but the mechanisms and drivers of this microbial attachment and colonization, as well as subsequent fermentation remain to be elucidated. We have previously shown that an individually dependent subset of gut bacteria is able to colonize the wheat bran residue. Here, we isolated these bran-attached microorganisms, which can then be used to gain mechanistic insights in future pure culture experiments. Four healthy fecal donors were screened to account for inter-individual differences in gut microbiota composition. A combination of a direct plating and enrichment method resulted in the isolation of a phylogenetically diverse set of species, belonging to the Bacteroidetes, Firmicutes, Proteobacteria and Actinobacteria phyla. A comparison with 16S rRNA gene sequences that were found enriched on wheat bran particles in previous studies, however, showed that the isolates do not yet cover the entire diversity of wheat-bran colonizing species, comprising among others a broad range of Prevotella, Bacteroides and Clostridium cluster XIVa species. We, therefore, suggest several modifications to the experiment set-up to further expand the array of isolated species.

Human urine is a valuable resource for nutrient recovery, given its high levels of nitrogen, phosphorus and potassium, but the compositional complexity of urine presents a challenge for an energy-efficient concentration and refinery of nutrients. In this study, a pilot installation combining precipitation, nitrification and electrodialysis (ED), designed for one person equivalent (1.2 L_urine d⁻¹), was continuously operated for ∼7 months. First, NaOH addition yielded calcium and magnesium precipitation, preventing scaling in ED. Second, a moving bed biofilm reactor oxidized organics, preventing downstream biofouling, and yielded complete nitrification on diluted urine (20–40%, i.e. dilution factors 5 and 2.5) at an average loading rate of 215 mg N L⁻¹ d⁻¹. Batch tests demonstrated the halotolerance of the nitrifying community, with nitrification rates not affected up to an electrical conductivity of 40 mS cm⁻¹ and gradually decreasing, yet ongoing, activity up to 96 mS cm⁻¹ at 18% of the maximum rate. Next-generation 16S rRNA gene amplicon sequencing revealed that switching from a synthetic influent to real urine induced a profound shift in microbial community and that the AOB community was dominated by halophilic species closely related to Nitrosomonas aestuarii and Nitrosomonas marina. Third, nitrate, phosphate and potassium in the filtered (0.1 μm) bioreactor effluent were concentrated by factors 4.3, 2.6 and 4.6, respectively, with ED. Doubling the urine concentration from 20% to 40% further increased the ED recovery efficiency by ∼10%. Batch experiments at pH 6, 7 and 8 indicated a more efficient phosphate transport to the concentrate at pH 7. The newly proposed three-stage strategy opens up opportunities for energy- and chemical-efficient nutrient recovery from urine. Precipitation and nitrification enabled the long-term continuous operation of ED on fresh urine requiring minimal maintenance, which has, to the best of our knowledge, never been achieved before.